PyCO2SYS is a Python implementation of CO2SYS, based on the MATLAB v2.0.5 but also including the updates made for MATLAB CO2SYS v3 as well as some additional related calculations. PyCO2SYS solves the full marine carbonate system from the values of any two of its parameters.
Every combination of input parameters has been tested, with differences in the results small enough to be attributable to floating point errors and iterative solver endpoint differences (i.e. negligible). See the scripts in validate to see how and check this for yourself, and their discussion in the online docs. Please let us know ASAP if you discover a discrepancy that we have not spotted!
Documentation is available online at PyCO2SYS.readthedocs.io.
There are also some usage examples that you can either download or run live in your web browser (with no Python installation required) at PyCO2SYS-examples.
The citation for PyCO2SYS alone is:
Humphreys, M. P., Gregor, L., Pierrot, D., van Heuven, S. M. A. C., Lewis, E. R., and Wallace, D. W. R. (2020). PyCO2SYS: marine carbonate system calculations in Python. Zenodo. doi:10.5281/zenodo.3744275.
The DOI above refers to all versions of PyCO2SYS. Please also specify the version number that you used. You can find this in Python with:
import PyCO2SYS as pyco2
pyco2.say_hello()As per the instructions in the the CO2SYS-MATLAB repo, you should also cite the original work by Lewis and Wallace (1998).
Install from the Python Package Index:
pip install PyCO2SYS
Update an existing installation:
pip install PyCO2SYS --upgrade --no-cache-dir
To solve the marine carbonate system from two of its parameters (par1 and par2), just use:
import PyCO2SYS as pyco2
results = pyco2.sys(par1, par2, par1_type, par2_type, **kwargs)Each field in the results corresponds to a column in the original MATLAB output DATA or to a new output that is only available from PyCO2SYS. The keys to the dict come from the original MATLAB output HEADERS. Inputs should be provided as scalars or NumPy arrays in any mutually broadcastable combination. A large number of optional kwargs can also be provided to specify everything beyond the carbonate system parameters - read the docs!.
There's also an alternative interface that is very similar to that of CO2SYS for MATLAB, which you can use instead if you prefer. It is also possible to provide Pandas Series and Xarray DataArrays using the wrapper functions provided. For this and a more detailed explanation of all the inputs and outputs, see the Pythonic API documentation.
You can also look at the example scripts here in the repo, or there are also some examples as Jupyter Notebooks that you can try out without needing to install anything on your computer.
PyCO2SYS is maintained by Dr Matthew Humphreys of the NIOZ (Royal Netherlands Institute for Sea Research) with support from the main developers of all previous versions of CO2SYS.
Contributions are welcome; please check the guidelines before setting to work.
PyCO2SYS is licensed under the GNU General Public License version 3 (GPLv3).